IP VERSION 6 (IPV6) Mario Baldi M. Baldi: see page 2. IPv6-1

Transcription

1 IP VERSION 6 (IPV6) Mario Baldi IPv6-1

2 IPv6-2 Copyright Notice This set of transparencies, hereinafter referred to as slides, is protected by copyright laws and provisions of International Treaties. The title and copyright regarding the slides (including, but not limited to, each and every image, photography, animation, video, audio, music and text) are property of the authors specified on page 1. The slides may be reproduced and used freely by research institutes, schools and Universities for non-profit, institutional purposes. In such cases, no authorization is requested. Any total or partial use or reproduction (including, but not limited to, reproduction on magnetic media, computer networks, and printed reproduction) is forbidden, unless explicitly authorized by the authors by means of written license. Information included in these slides is deemed as accurate at the date of publication. Such information is supplied for merely educational purposes and may not be used in designing systems, products, networks, etc. In any case, these slides are subject to changes without any previous notice. The authors do not assume any responsibility for the contents of these slides (including, but not limited to, accuracy, completeness, enforceability, updated-ness of information hereinafter provided). In any case, accordance with information hereinafter included must not be declared. In any case, this copyright notice must never be removed and must be reported even in partial uses.

3 Outline A new version of IP: why? Addresses Modified protocols Socket programming interface Packet header format Neighbor discovery Transition to IPv6(?) IPv6-3

4 A NEW VERSION OF IP: WHY? IPv6-4

5 Why a new IP? Only one true answer A larger address space IPv6-5

6 Other answers More efficient on LANs Multicast and anycast Security Policy routing Plug and play Traffic differentiation Mobility Quality of service support Ported to IPv4 IPv6-6

7 A long way to IPv6 adoption Long time for defining IPv6 and migrating to it Problems needed an interim solution in IPv4 When IPv6 reached "production" stage, many IPv4 solutions were acceptable IPv6-7

8 Why are IPv4 addresses scarce? 32 bit long About 4 billion addresses!!! however... IPv6-8

9 Only part of the addresses are assigned to stations Class A, B and C Addresses beginning by bx111 are used for multicast and else Hence, just 3.5 billion addresses can be used!!! IPv6-9

10 They are used hierarchically The prefix used in a physical network cannot be used in a different one Lots of unused addresses IPv6-10

11 Interim (IPv4) solutions to the saturation of address space Introduction of network with "taylored" size Netmask Private addresses Intranet, RFC 1918 Not enough to solve the problem It should be used in conjunction with NAT or ALG Network Address Translator (NAT) Extremely popular Proposal for RSIP (Realm Specific IP) ALG (Application Level Gateway) IPv6-11

12 Has all of this been effective? IPv6-12

13 Agencies assigning addresses IANA distributes (better: distributed) /8 IPv4 network prefixes to regional registries IPv6-13

14 Situation (2010) IPv6-14

15 Situation (2011) IPv6-15

16 IPv6-16

17 IPv6-17

18 Routing scalability issues Routing table size Internet size Each subnetwork must be advertised Problems Router resource limitations Too much information to manage Routing protocol limitations High probability of route changes Mainly affecting backbone routers IPv6-18

19 Routing scalability issues IPv6-19

20 Isn t there a solution with IPv4? Aggregate multiple routes in one Shorter prefix including others /24, / /16 CIDR (Classless Inter-Domain Routing) Limited by non-rational assignment of IP prefixes IPv6-20

21 IPv6-21 Interim (IPv4) solutions to routing scalability CIDR Classless Inter-Domain Routing Limiting the assignment of IP addresses Regional Internet Registry: assign address blocks only to big players E.g., minimum /20 (4096 addresses) network Scalability of routing protocols With no solution, at present Problem not completely solved It is the major problem that IPv6 wanted to solve that it is still open

22 IPv6-22 Birth of IPv6 IETF Boston Meeting (1992), Call for proposals Appointment of dedicated Working Groups Several proposals TUBA: adopting OSI CLNP as new IP CATNIP: integration of different network (IP, CLNP, IPX) and transport (TP4, SPX, TCP, UDP) protocols SIPP: incremental over IPv4 Fix some drawbacks Simple: increasing the address field and eliminating unused ones Winning proposal: SIPP with 128 bit addresses

23 ADDRESSES IPv6-23

24 So, how many addresses should IPv6 have? A scientific approach Addressing efficiency H = log 10 (number of addresses) number of bits IPv6-24

25 Addressing Efficiency In existing networks H varies between O.22 and O.26 Assuming one million billion networked stations 68 bits in the minimum efficiency case IPv6-25

26 Melius abundare quam deficere 128 bits (16 bytes) O IPv6 addresses per sqm of Hearth surface IPv6-26

27 Notation 8 hexadecimal numbers separated by : Groups of 2 bytes FEDC:BA98:0876:45FA:0562:CDAF:3DAF:BB :0000:0000:0007:0200:A00C:3423:A089 IPv6-27

28 Shortcuts Leading Os in each digit group can be omitted 1O8O:O:O:O:7:2OO:AOOC:3423 Groups of Os can be substituted by :: 1O8O::7:2OO:AOOC:3423 Not more than once IPv6-28

29 Addressing Space Organization Multicast FF00::/8 FFxx:xx FFFF:.:FFFF FF00::0 Multicast 0::0 IPv6-29

30 HOST ADDRESSES Routing and Addressing Principles IPv6-30

31 Same routing principles as IPv4 Subnet 4 Subnetwork 1 Subnetwork 2 Subnet 3 Router IPv6-31

32 Address Structure n bit Prefix 128 n bit Interface Identifier n = 64 IPv6-32

33 Same Address Assignment Principles as IPv4 (different terminology) Sub network: set of hosts with same prefix Link: physical network Subnetwork link IPv6-33

34 On-link hosts have same prefix Communicate directly Off-link stations have different prefix Communicate through a router IPv6-34

35 Prefix Address/netmask pair is substituted by a Prefix Address/N, where N is the prefix length [bit] FEDC:0123:8700::/ No address classes IPv6-35

36 Link local/site local O 1 Link local O 1O FE8O::/64 Site local (deprecated) O 11 FEC0::/1O FE[C-F] Equivalent to IPv4 private addresses FFFF:.:FFFF FF0::0 FE80::0 0::0 Site local Multicast Link/Site local IPv6-36

37 Why Deprecated? Overlapping private address spaces Not a problem in principle, but in practice Extranets Mergers and acquisitions IPv6-37

38 Private Addresses Unique Local Addresses (ULA) FC00::/ FC00::/8 for future use FFFF:.:FFFF FFFF::0 FE80::0 FDFF:FFFF:.:FFFF FD00::0 FC00::0 Multicast Link/Site local Private Other ULA Private addresses FD00::/8 0::0 IPv6-38

39 Local Unicast Addresses private randomly generated Subnet ID Interface ID FD site local FE[C-F]x any Interface ID link local Interface ID FE80 IPv6-39

40 Remaining addresses Global Unicast IPv6-40

41 Global Unicast Addressing Space Organization FFFF:.:FFFF Multicast IPv4 interoperability (80 bit) 0::/80 To be used during transition phase IPv4 mapped addresses 16 bits set to 1 0:0:0:0:0:FFFF::/96 FFFF::0 FE80::0 FDFF:FFFF:.:FFFF FD00::0 FC00::0 0::FFFF:FFFF:FFFF 0::0 Link/Site local Private Other ULA IPv4 interoperability IPv6-41

42 IPv4 compatible Another 16 bits to 0 0::/96 E.g. 0:0:0:0:0:0:A00:1 Compact notation ::A00:1 Special notation :: IPv6-42

43 Aggregatable Global Unicast Begin with bxoo1 [2-3] Topology-based assignment Service provider hierarchy Effective aggregation FFFF:.:FFFF FFFF::0 FE80::0 FDFF:FFFF:.:FFFF FD00::0 FC00::0 3FFF:.:FFFF 2000::0 0::FFFF:FFFF:FFFF 0::0 Multicast Link/Site local Private Other ULA Aggregatable Global Unicast IPv6-43

44 Tier 2 ISP Tier 1 ISP Tier 3 ISP Tier 3 Tier 2 ISP ISP Tier 2 ISP Tier 1 ISP Tier 3 ISP Tier 3 ISP Tier 1 ISP Tier 2 ISP Tier 2 ISP Tier 3 ISP Tier 2 ISP Tier 3 ISP Different assignment criterion for other addresses IPv6-44

45 Address Assignment bit bit 16 bit TLA ID Prefix NLA ID SLA ID Top level authority Large ISP Next level authority Organization 64 bit Interface Identifier Ethernet Address (EUI 64) Subnet level authority Automatic renumbering IPv6-45

46 More Flexibility 48 64bit bit 16 bit 64 bit Global Routing Prefix Prefix Subnet ID Interface Identifier bit bit Global Routing Prefix Prefix 16 bit Subnet ID 64 bit Interface Identifier 2001::/ ::/23 delegated to registries IPv6-46

47 Plug and Play Scenarios Dentist Office Thousand computers on the dock Solution: autoconfiguration Stateless: no server needed Statefull: DHCP server IPv6-47

48 Special Addresses Loopback address ::1 All nodes (multicast) address FF02::1 All routers (multicast) address FF02::2 Unspecified address :: IPv6-48

49 MODIFIED PROTOCOLS IPv6-49

50 What changes in the protocol architecture? IP ICMP ARP Integrated in ICMP IGMP Integrated in ICMP IPv6-50

51 Upgraded But Not Changed DNS (type AAAA record) RIP and OSPF BGP and IDRP TCP and UDP Socket interface What about layer independence? IPv6-51

52 SOCKET PROGRAMMING INTERFACE IPv6-52

53 What is it? Programming interface for TCP/IP services Used in application implementation UDP messages Bytes on TCP connections IPv6-53

54 Underlying Principles Originated in Unix Environment I/O as file access Socket descriptor equivalent to a file descriptor for network use IPv6-54

55 Socket Point of access to network services Associated to TCP connection or UDP session IPv6-55

56 Socket Operations Wait for connection requests on a port Server listen() Accept requests (server) IPv6-56

57 Connect to a port of a remote server Client Requires specifying address and port Send and receive data IPv6-57

58 PACKET HEADER FORMAT IPv6-58

59 Do You Remember the IPv4 Header? VER HLEN ToS Total Length Identifier Flag Source Address Destination Address Fragment Offset TTL Protocol Checksum Options PAD IPv6-59

60 Here is the IPv6 One 40 bytes VER Priority Payload length Next header Source Address Destination Address Flow label Hop limit Simple and constant length IPv6-60

61 Field Removal Not very useful Checksum Not used in each packet Fragmentation No longer needed Header length IPv6-61

62 Extension Headers Added when useful Not needlessly processed in each packet IPv6-62

63 Extension Headers Hop By Hop Option Routing Fragment Authentication Encrypted Security Payload Destination Option IPv6-63

64 Extension Header Format Next Header Length Extension data More extension data More extension data IPv6-64

65 Header Chaining IPv6 Header N.H.=TCP TCP Segment IPv6 Header N.H.=Routing Routing header N.H.=TCP TCP Segment IPv6 Header N.H.=Routing Routing header N.H.=Fragm. Fragm. header N.H.=TCP TCP Segment IPv6-65

66 Options To be used in Hop-by-hop and Destination Option Extension Headers TLV format Type - Length Value Type Length Value IPv6-66

67 Sample Option Usage Next Header Length Type 1 4 Value Type 2 6 Value Value IPv6-67

68 Sample Option: Jumbo Payload Type Length Value Jumbo Payload Length Jumbo Payload Length IPv6-68

69 Padding Options Extension Headers must be 64 bit aligned PadN Option Type Pad1 Option Type 0 Length Value bytes of data IPv6-69

70 Type Field: first three bits First 2 bits: action in case the option is not recognized Code Meaning 00 The current option can ignored. It is possible to proceed with the next one 01 The packet must be discarded 10 The packet must be discarded and an ICMPv6 Parameter Problem message generated 11 The packet must be discarded and an ICMPv6 Parameter Problem must be generated, unless the destination address is a multicast one Third bit: indicates if the option can be modified on-the-fly Code Meaning 0 The option cannot be changed on-the-fly 1 The option can be changed on-the-fly IPv6-70

71 Routing Header Used by an IPv6 source to list nodes to traverse on the path to the destination It can be a loose route Segment Left Field shows the number of the remaining path segments Points to the next router to reach IPv6-71

72 Next Header Reserved Hdr Ext Len Routing Type Segments Left Strict/Loose Bit Map Address[1] Address[2] Address[n] IPv6-72

73 Routing Header: example S R1 R2 D IPv6 Hdr From: S To: R1 NextHdr: Routing Routing Hdr Segment Left: 2 Hop 1: R2 Hop 2: D IPv6 Hdr From: S To: R2 NextHdr: Routing Routing Hdr Segment Left: 1 Hop 1: R1 Hop 2: D IPv6 Hdr From: S To: D NextHdr: Routing Routing Hdr Segment Left: 0 Hop 1: R1 Hop 2: R2 IPv6-73

74 INTERFACING WITH THE LOWER LAYER IPv6-74

75 Encapsulation Encapsulated in layer 2 frames EtherType: 86DD As a new protocol Enables dual stack approach Keep running IPv4 as-is IPv6-75

76 Address Mapping What is the destination MAC address? IP unicast address Procedural (protocol-based) discovery Neighbor Discovery IP multicast address Algorithmic mapping IPv6-76

77 IPv6 Multicast Transmission Based on MAC multicast IPv6 multicast address mapped to MAC address least significant bytes of IPv6 address 127 FF Last 32 bits of the IPv6 address Last 32 bit of the IPv6 address 0 0 IPv6-77

78 Multicast Address Mapping Example When sending a packet to the IP multicast address FFOC::89:AABB:CCDD Encapsulate in MAC frame to 33:33:AA:BB:CC:DD IPv6-78

79 NEIGHBOR DISCOVERY IPv6-79

80 New Function in ICMP It substitutes ARP Based on multicast Most likely only one station gets involved IPv6-80

81 Solicited Node Multicast Address Subscribed by all hosts FFO2::1:FF/1O4 24 least significant bits of IP address Likely 1 host per group IPv6-81

82 Address Resolution ICMP Neighbor Solicitation To Solicited Node Multicast Address of target IPv6 address ICMP Neighbor Advertisement To requester address IPv6-82

83 Resolution Example To find the MAC address of host 2OO1::ABCD:EF98 ICMP Neigh Sol to Sol Node Mult Add: FFO2::1:FFCD:EF98 Encapsulate in MAC frame to 33:33:FF:CD:EF:98 IPv6-83

84 Host Cache Mapping between IPv6 and MAC address Equivalent to ARP cache IPv6-84

85 TRANSITION TO IPv6 (?) IPv6-85

86 IPv4 to IPv6 Transition Incremental Seamless Smooth IPv6-86

87 How can we enable this? Dual-stack approach IPv6 as a new protocol Generate/receive v6 or v4 packets as needed Address mapping Tunneling Translation mechanisms Application IPv4 IPv6 Ethernet IPv6-87

88 Dual stack Hosts Isolated IPv6 Networks IPv4 IPv6 in IPv4 Tunnel IPv6 IPv6 IPv6 IPv6 IPv6-88

89 IPv6 Islands Grow IPv6 only Hosts IPv4 IPv6 IPv6 IPv6 IPv6 Dual stack Translating Devices IPv6 IPv6 IPv6 IPv6 IPv6 IPv6-89

90 Native IPv6 Connectivity IPv6 only Hosts IPv4 IPv6 IPv6 IPv6 IPv6 Dual stack Translating Devices IPv6 IPv6 IPv6 IPv6 IPv6 IPv6-90

91 All the Way to the Doomsday IPv6 IPv4 IPv4 IPv6 IPv4 IPv6 IPv4 in IPv6 Tunnel IPv4 IPv6 IPv6 IPv4 IPv6 IPv6-91

92 Are we ready? All protocols specified For a while: since 1996!! IPv6-92

93 Implemented on routers Even if less stable than IPv4 Possibly not all functionalities Some hardware implementations (Layer 3 switch) IPv6-93

94 Implemented in end systems Windows since 2OOO and XP Unix, FreeBSD, Linux Quite a few applications Possibly with a few bugs IPv6-94

95 When will it happen? Large IPv4 install base Only one true motivation: Address space depletion IPv6-95

96 The issue has been mitigated Provident address assignment Extensive use of private addressing NAT and proxying IPv6-96

97 So, don t we need IPv6? NAT not suitable for all applications Problematic with security mechanisms IPv6-97

98 User traceability Not practical with servers Not many pubblic addresses Acceptable limitations so far IPv6-98

99 Just Plain Address Space Exhaustion Especially in the Asia-Pacific region IANA ran out of class A prefixes in Feb 2O11 RIPE by end 2O11 Possibly legislation IPv6-99

100 165OO Allocated Prefixes Allocated prefixes Announced Alive June 2003 June 2013 IPv6-100

101 Current IPv6 web deployment IPv6-101